This investigation will focus on two key brain regions suggested by recent studies to be vulnerable in people at risk for Alzheimer Disease (AD)-the mesial temporal lobe, and the posterior cingulate cortex (PC). The role of the mesial temporal lobe in forming new memories has been well established with lesion studies and neuropsychological investigation. The role of the PC is only beginning to be understood (through functional imaging) as important in retrieval processes. Recent imaging studies suggest that both of these regions are affected in AD even before clinical symptoms appear, consistent with the very early memory symptoms in AD. In this project we will study middle age and older adults with one or more risk factors for AD including first-degree family history, ApoE genotype, or memory difficulty. Group one will be selected based on family history and genetic risk for AD (presence or absence of at least one APOE e4 allele) compared with age-matched controls. A second risk group will be patients identified with Mild Cognitive Impairment-amnestic type (MeIa) (and age matched controls), who will be enrolled over the first two years, imaged at baseline and monitored with cognitive testing until conversion to AD or until the end of the funding period. Half of the MCI patients are expected to convert over the five years of support. Aims: Using high-field (3Tesla) functional MRI (fMRI) and extensive clinical and cognitive characterization, our aims are 1) To determine the independent contributions of age, APOE status, and presence or absence of family history of AD, to the cerebral response in the mesial temporal lobe and posterior cingulate during learning and retrieval tasks in cognitively healthy adults age 46 to 65. 2) To show that patients with MCI differ from age-matched controls in the MTL and PC response, and that conversion to incident AD is predicted by the baseline PC and MTL fMRI response. Methods: Voxel-based image processing methods will be applied to regions of interest for the fMRI data using clinical characteristics as grouping variables and covariates within the general linear model to accomplish the above aims. Significance: Examining learning and retrieval functions in asymptomatic and symptomatic persons with risk factors for AD may provide rich insight into the organization of learning and memory systems and how they may begin to fail in preclinical AD. In the future this information may lead to improved methods for earlier detection of the disease.